Can it handle multiple transcripts at once?

Yes, you can use glob patterns to process multiple files. The skill will then perform a cross-transcript analysis to find duplicates or conflicting statements.

What file formats does the analyze-transcript skill support?

The skill supports text-based files including .md (Markdown) and .txt files containing meeting notes, transcripts, or conversational logs.

Where does the skill save the extracted information?

Results are saved into a structured .requirements directory, organized by domain and transcript name in YAML format for easy version control and reading.

How does the skill prioritize different speakers?

It uses a weighting system where roles like Product Owners and Technical Leads are given higher priority for business and technical requirements respectively, while unknown speakers are flagged for validation.

Meeting Transcript Requirement Extractor

Name: Meeting Transcript Requirement Extractor
Author: melodic-software

bymelodic-software

•

コラボレーションツール

Extracts requirements, decisions, and action items from meeting transcripts using specialized conversational pattern recognition.

The analyze-transcript skill automates the transition from raw meeting notes to structured project data. By processing conversational content from stakeholder calls, design reviews, or sprint planning, it identifies explicit and implicit requirements, technical decisions, and pending action items. It features a unique speaker-weighting system that prioritizes input from key roles like Product Owners and Tech Leads, while also detecting conflicts across multiple transcript files to ensure project alignment and documentation accuracy.

主な機能

01Speaker role identification with importance-based weighting

0238 GitHub stars

03Automated identification of risks, concerns, and action items

04Cross-transcript analysis to detect conflicting requirements

05Structured YAML output for seamless integration with requirement management

06Specialized extraction for conversational markers and decision patterns

ユースケース

01Converting stakeholder interview transcripts into a prioritized product backlog

02Consolidating multiple project discovery calls into a unified requirement document

03Capturing technical architecture decisions and rationales from design reviews

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add melodic-software/claude-code-plugins analyze-transcript

For use in Claude.ai and ChatGPT

Download Skill